The purpose of the proposed research is to develop, for use in epidemiologic studies, a method for estimating the degree of human exposure to organic chemical contaminants in drinking water. The study will, in its first phase: 1) monitor the levels of four trihalomethanes (THMs) in the Columbus, Ohio, water supply system; and 2) use these chemical determinations and routinely collected water treatment plant data to construct a predictive mathematical-statistical model of THM formation. The model will be developed to permit the estimation of THM levels in past time periods when the only information available was the routine water plant data. In the second phase of the study, the predictive model will be applied to generate estimated HTM levels for Columbus for the period 1956-1980. From these estimates a quantitative index of human exposure to waterborne THMs will be derived. Procedures for applying the model to other water systems and to epidemiologic studies will be specified.